Vitamin D can have powerful anti-carcinogenic effects and oral supplementation offers the potential to reduce breast cancer risk in the US population, where a large proportion of women have insufficient levels. In a recent prospective cohort study, we showed that high levels of serum 25-hydroxyvitamin D [25(OH)D] were associated with decreased risk of breast cancer over the subsequent five years. We seek to expand upon our existing work by investigating how 25(OH)D levels change over time and the implications of these changes for assessing the association between vitamin D and breast cancer. We will do this by examining how breast cancer diagnoses and other time-varying, health-related factors affect changes in women?s 25(OH)D levels. To conduct this work, we will use data from the Sister Study, a large prospective cohort study of women who had a sister with breast cancer but had never had breast cancer themselves when they enrolled (2003-2009). We have already assessed baseline 25(OH)D levels in 3,386 participants (including 1,611 women who developed breast cancer and 1,775 women who did not). Serum specimens were collected from 2,430 of these same women in 2013. In this proposal, we will measure 25(OH)D levels in the repeat samples.